Abstract
Aims:
To assess whether there is an opportune window when intensive lifestyle intervention (ILI) benefits cognitive function
Methods:
Standardized cognitive assessments were collected following ≥8 years of either ILI or a control condition of diabetes support and education (DSE) in 3,708 individuals, ages 45–76 years at enrollment, with type 2 diabetes and overweight or obesity. Frailty index (FI) scores were used to group individuals at baseline into tertiles according to their age-related health status. Linear models were used to describe intervention adherence and cognitive function, with interaction terms to examine the consistency of relationships among tertiles.
Results:
Worse baseline FI scores were associated with poorer subsequent performance in tests of attention, processing speed, and executive function. No differences in any measure of cognitive function were observed between intervention groups within any FI tertile (all p>0.10). Among individuals with worse baseline FI scores, weight gain was associated with poorer global cognitive function among participants assigned to DSE. There was no association between weight changes and cognitive function among participants assigned to ILI.
Conclusions:
Among adults with type 2 diabetes and overweight/obesity, we found no evidence that there is a window of opportunity based on FI when ILI benefits cognitive function.
Keywords: Intensive lifestyle intervention, cognitive function, aging, type 2 diabetes mellitus, obesity
1. Introduction
Type 2 diabetes and obesity, separately and in combination, accelerate cognitive decline, increasing risk for dementia.1,2 While lifestyle intervention including weight loss may be expected to mitigate against this through multiple pathways, including improved diabetes and blood pressure control, reduced inflammation, and reduced sleep apnea, the Action for Health in Diabetes randomized controlled clinical trial (Look AHEAD) found no overall cognitive benefit of an intensive lifestyle intervention (ILI) compared with a control condition of diabetes support and education (DSE).3,4 However, there was evidence that associations between intervention assignment and cognitive function were heterogeneous. These associations ranged from potential benefit for individuals with lower body mass index (BMI) and no history of cardiovascular disease, to potential harm for the heaviest individuals with cardiovascular disease history. Significant interactions between BMI and/or history of cardiovascular disease with intervention assignment were observed for objective measures of cognitive function,3–5 cognitive impairment6 and self-reported cognitive ability.7
We conducted an exploratory analysis to examine whether there was differential response to the ILI with respect to the Look AHEAD primary outcome, a composite of myocardial infarction, stroke, death from cardiovascular disease, and hospitalization for angina.8 For this outcome, ILI appeared only to be beneficial for those with relatively better baseline health profiles as captured by a frailty index (FI) based on deficit accumulation.9.10 We interpreted this as evidence that there may be an opportune window based on an individual’s underlying health status when lifestyle intervention may be most effective in reducing cardiovascular disease risk. Because cardiovascular disease and cognitive decline share many risk factors, this led us to conjecture that the relative impact of ILI on cognitive function may also vary depending on baseline FI.
There continue to be concerns about the impact of weight loss intervention on cognitive decline in older individuals, often related to the observation that weight loss is associated with increased risk for cognitive impairment in older individuals.11 This association is often attributed to underlying pathologies that may lead to unintentional weight loss, however there remains a possibility that features of the underlying health status may influence how weight loss affects cognitive function. Successful trials supporting a beneficial association between weight loss and cognitive function in older individuals alone or in combination with physical activity, e.g.12,13, have been relatively short term and small, and underpowered to assess differences in response among subgroups based on health status, and they had limited participation of people with diabetes. However, benefits were observed to continue past the termination of the intervention The largest trials of long term effects of intentional weight loss on individuals at risk for diabetes were both null and only examined the consistency of intervention effects in limited numbers of subgroups.14,15
2. Methods
The Look AHEAD protocol and CONSORT diagram have been published.16,17 It was a multi-site, single-masked randomized controlled clinical trial that recruited 5,145 individuals (during 2001 to 2004) from 16 U.S. centers. All had type 2 diabetes and met the following criteria: 45–76 years of age, BMI >25 kg/m2 (>27 kg/m2 if on insulin), glycated hemoglobin (HbA1c) <97mmol/mol (11%), systolic/diastolic blood pressure <160/<100 mmHg, triglycerides <600 mg/dl, a successful maximum graded exercise test. Protocols and consent forms were approved by local Institutional Review Boards.
2.1. Interventions
Participants were randomly assigned to ILI or DSE. The ILI targeted reducing caloric intake and increasing physical activity to induce weight loss >7% and maintaining this over time.18,19 Caloric consumption goals of 1200–1800 kilocalories/day were based on initial weight. Physical activity of >175 minutes/week through activities similar in intensity to brisk walking was targeted, as was improved diet (<30% calories from fat, <10% calories from saturated fat, >15% calories from protein). Cardiometabolic risk factors (lipids, HbA1c, blood pressure) were monitored and participants were provided results. During the first six months, ILI participants attended three group meetings and one individual session per month. For the remainder of the first year, they were provided two group and one individual meeting per month. The intensity of the intervention gradually decreased thereafter.18
DSE participants were invited to attend group sessions focused on diet, physical activity, and social support.19 Four meetings were offered during year 1, three per year during years 2–4, and one annually thereafter. Participants did not receive specific diet, activity, or weight goals or information on behavioral strategies, however the protocol for sharing risk factor information with participants and their physicians was the same as for ILI.
Interventions ended September 2012, when participants’ planned follow-up ranged from 8–11 years.
2.2. Intervention response
Measures were obtained by staff masked to intervention assignment.16 Weights were measured annually. A maximal graded exercise test was administered at baseline and a submaximal test at years 1 and 4.20 Changes in fitness were computed as differences between estimated metabolic equivalents (METS) when participants achieved or exceeded 80% of age-predicted maximal heart rate or Borg Rating of Perceived Exertion of >16, at baseline and subsequently.
2.3. Cognitive function
Cognitive assessment was performed among those enrolled in a post-intervention observation study between August 2013 and December 2014, 10–13 years after their Look AHEAD enrollment. A subset had one or two earlier assessments in the Look AHEAD Movement and Memory Study (4 clinics: years 8–11)3 and the Look AHEAD Brain MRI Study (3 clinics: years 10–12).21 We used data from the last assessment if participants had more than one.
Assessments were performed by centrally trained, masked staff.3 Verbal learning and memory were evaluated with the Rey Auditory Verbal Learning Test (RAVLT). Speed of processing and working memory were evaluated with the Digit Symbol Coding test (DSC). Executive function was evaluated with the Modified Stroop Color and Word Test (MSCWT) and the Trail Making Test-Part B (TMT-B). Global cognitive functioning was evaluated with the Modified Mini-Mental Status Exam (3MSE). Scores from individual tests were z-transformed, ordered so higher scores reflected better performance, and averaged to form domains.3 A composite cognitive function score was the average z-transformed scores from individual tests, which was pre-specified as the primary cognitive endpoint for the Look AHEAD trial.22
2.4. Deficit accumulation FI
We constructed a 38-item FI modeled after one developed by the Systolic Blood Pressure Intervention Trial,10 which we augmented with nine additional deficits related to diabetes and obesity (Supplemental Exhibit S1).23 Included were measures of behavior, medical history, clinical measures, function, and abilities. Individual component scores range between 0 and 1, with higher scores represening greater frailty. The total FI ratio is the sum of the individual component scores divided by the number of components, ranging from 0 to 1.
2.5. Baseline risk factors
Blood pressure was measured in duplicate. Blood specimens were collected after ≥12-hour fasting and analyzed centrally. Self-reported characteristics were assessed with questionnaires and interviews. CVD history was self-report of prior myocardial infarction, coronary artery bypass, angioplasty/stent procedures, peripheral vascular disease, stroke, stable angina, and class I /II heart failure. Hypertension was current treatment or blood pressure >140/90 mmHg. Depressive symptoms were assessed with the Beck Depression Index.
2.6. Statistical analysis
We analyzed de-identified data developed for investigators outside the Look AHEAD study group. Of the 5,145 participants, 4,901 (95.3%) provided consent for data sharing; data were sufficient to compute baseline FI scores for 4,859 (99.1%). Cognitive assessments were available from N=3,708 (76%) of these participants. Because there was evidence that lost follow-up varied among participant subgroups, which may introduce bias, we used inverse probability weighting to gauge the influence of differential attrition on findings. Logistic regression was used to model the probability of attrition according to baseline characteristics (markers of health, demography, and behavior and intervention assignment), and inverses of the modeled probabilities were used to weight inferences.24 We assessed the sensitivity of this approach by comparing results from two models with varying numbers of predictors of attrition and found no material differences.
Baseline characteristics among FI tertile groups were compared using chi-squared tests and analyses of variance. We examined differences between the participants contributing to the current analysis from those who were not included (i.e., received no cognitive assessments) using chi-squared and t-tests.
We used analyses of covariance to examine mean differences in standardized cognitive assessments scores among FI tertiles within each intervention group, with covariate adjustment for age, sex, education, race/ethnicity, and the years between randomization and the time of the cognitive assessment. Interaction terms were included to assess the consistency of differences in scores between intervention groups across FI tertiles and linear contrasts were used to estimate mean differences in scores between intervention groups within each tertile.
The correlations between 8-year changes in body mass index and the composite cognitive function scores were calculated for FI tertile groups within each intervention. Linear regression with interaction terms was used to assess the consistency of these relationships across tertiles and intervention groups.
3. Results
FI scores at baseline ranged from 0.066 to 0.588, with median 0.202. The upper boundaries of the first and second tertiles were 0.178 and 0.230, respectively. Table 1 describes baseline characteristics by FI tertile. Those in the highest tertile were more likely to be women and African-Americans, and to have lower education and no history of CVD. They tended to have higher systolic blood pressure, higher BMI, longer diabetes durations, and poorer metabolic profiles. Intervention assignment and time between enrollment and cognitive assessments were balanced across FI tertiles. Mean FI scores did not differ significantly between age groups and was only weakly correlated (r=0.11) across the full age range (Supplement S2).
Table 1:
Characteristics at Look AHEAD enrollment by tertile of the Rockwood Deficit Accumulation Frailty Index: N (percent) or mean (standard error).
| Deficit Accumulation Frailty Index | ||||
|---|---|---|---|---|
| First Tertile | Second Tertile | Third Tertile | ||
| N= 1306 | N=1253 | N=1149 | p-value* | |
| Range | Range | Range | ||
| [0.066,0.178) | [0.178, 0.230) | [0.230,0.588] | ||
| Age, years | ||||
| 45–59 | 781 (59.80) | 723 (57.70) | 668 (58.14) | 0.52 |
| 60–76 | 525 (40.20) | 530 (42.30) | 481 (41.86) | |
| Sex | ||||
| Female | 687 (52.60) | 780 (62.25) | 760 (66.14) | <0.01 |
| Male | 619 (47.40) | 473 (37.75) | 389 (33.86) | |
| Race/Ethnicity | ||||
| African-American | 204 (15.62) | 212 (16.92) | 231 (20.10) | |
| Hispanic | 185 (14.17) | 155 (12.37) | 172 (14.97) | 0.02 |
| Non-Hispanic White | 870 (66.62) | 839 (66.96) | 699 (60.84) | |
| Other, multiple | 47 (3.60) | 47 (3.75) | 47 (4.09) | |
| Education, miss=96 | ||||
| <13 years | 223 (17.49) | 218 (17.75) | 249 (22.29) | |
| 13–16 years | 412 (32.31) | 457 (37.22) | 488 (43.69) | <.0.01 |
| >16 years | 640 (50.20) | 553 (45.03) | 380 (34.02) | |
| BMI, kg/m2 | 34.04 (5.14) | 35.95 (5.68) | 37.72 (6.23) | <0.01 |
| CVD Risk factors | ||||
| Systolic blood pressure, mmHg | 125.51 (15.12) | 128.72 (16.91) | 131.38 (18.31) | <0.01 |
| Diastolic blood pressure, mmHg | 70.20 (9.06) | 70.21 (9.52) | 70.22 (10.00) | 0.99 |
| Total cholesterol, mg/dl, miss=8 | 190.91 (32.09) | 189.89 (38.28) | 193.18 (41.67) | <0.01 |
| HDL-cholesterol, mg/dl, miss=8 | 43.55 (11.39) | 43.56 (11.58) | 43.81 (12.73) | 0.06 |
| LDL-cholesterol, mg/dl, miss=9 | 113.86 (28.67) | 112.23 (32.51) | 111.52 (34.64) | <0.01 |
| Triglycerides, mg/dl, miss=8 | 169.83 (96.76) | 175.19 (107.34) | 195.10 (131.34) | <0.01 |
| HbA1c, %, miss=8 | 7.02 (0.98) | 7.24 (1.15) | 7.47 (1.27) | <0.01 |
| Diabetes duration, yrs, miss=23 | 5.42 (5.39) | 6.49 (6.14) | 7.72 (6.99) | <0.01 |
| History of cardiovascular disease | ||||
| No | 1230 (94.18) | 1103 (88.03) | 929 (80.85) | <0.01 |
| Yes | 76 (5.82) | 150 (11.97) | 220 (19.15) | |
| Intervention assignment | ||||
| DSE | 643 (49.23) | 639 (51.00) | 560 (48.74) | 0.50 |
| ILI | 663 (50.77) | 614 (49.00) | 589 (51.26) | |
| Years between randomization and cognitive assessment | ||||
| DSE | 11.27 (0.92) | 11.18 (0.97) | 11.09 (1.10) | <0.01 |
| ILI | 11.23 (0.95) | 11.15 (1.01) | 11.09 (0.98) | 0.60 |
Chi-squared or analysis of variance
As seen in Table 2, ILI participants, compared with DSE particpants, had larger mean changes in BMI over 8 years and relatively greater improvements in fitness measured from graded exercise tests over 4 years (the times when these were administered). Overall, these differences were evident for participants in each FI tertile.
Table 2:
Mean (standard error) percent changes from baseline in body mass index at years 1, 4, and the time of cognitive assessment and changes in fitness at years 1 and 3, by FI tertile. Mixed effects models were used to summarize overall differences across the three time points. Percent change in body mass index is calculated from the current difference from baseline divided by the baseline value.
| Year 1 | Year 4 | Time of Cognitive Assessment | Mean Difference Over Time p-value | |
|---|---|---|---|---|
| BMI Percent Change | ||||
| Overall Mean (SE) | ||||
| DSE | −0.72 (0.18) | −0.70 (0.18) | −1.43 (0.18) | |
| ILI | −8.92 (0.18) | −4.27 (0.18) | −4.11 (0.18) | |
| Difference (DSE-ILI) [95% CI] | ||||
| Overall | 8.21 [7.72,8.70] | 3.57 [3.08,4.06] | 2.68 [2.20,3.17] | <0.01 |
| First textile | 8.78 [8.01,9.55] | 3.90 [3.13,4.67] | 3.47 [2.70, 4.26] | <0.01 |
| Second textile | 8.28 [7.44,9.11] | 3.93 [3.09,4.76] | 2.23 [1.40, 3.07] | <0.01 |
| Third tertile | 7.49 [6.54,8.43] | 2.82 [1.88,3.77] | 2.30 [1.36, 3.25] | <0.01 |
| Fitness Change (METS) | Not Measured | |||
| Overall Mean (SE) | ||||
| DSE | 6.52 (0.64) | 0.13 (0.65) | ||
| ILI | 21.74 (0.63) | 5.69 (0.64) | ||
| Difference (DSE-ILI) [95% CI] | ||||
| Overall | −15.22 [−16.98,−13.46] | −5.56 [−7.35,−3.77] | <0.01 | |
| First tertile | −17.32 [−20.23,−14.40] | −6.65 [−9.60,−3.71] | <0.01 | |
| Second tertile | −14.42 [−17.51,−11.33] | −5.94 [−9.09,−2.79] | <0.01 | |
| Third tertile | −13.64 [−16.82,−10.45] | −4.05 [−7.29,−0.80] | <0.01 |
Composite cognitive function scores were inversely associated with baseline FI, both among ILI and DSE participants, with significant mean differences among tertiles, p<0.001 (ILI) and p=0.008 (DSE) respectively (Table 3). There was little evidence for differences in mean scores between intervention groups in any tertile and the interaction between intervention assignment and FI tertile was not significant (p=0.27).
Table 3:
Mean (SE) Cognitive function scores by intervention assignment and baseline deficit accumulation FI.
| Cognitive Outcome z-score | First Tertile of FI | Second Tertile of FI | Third Tertile of FI | Differences Across Tertiles Within Intervention Group P-value | Consistency of Intervention Group Differences Across Tertiles p-value |
|---|---|---|---|---|---|
| Composite | |||||
| ILI | −0.08 (0.03) | −0.19 (0.03) | −0.24 (0.03) | <0.001 | 0.27 |
| DSE | −0.12 (0.03) | −0.14 (0.03) | −0.23 (0.03) | 0.008 | |
| Mean Difference | |||||
| [95% CI] | 0.03 [−0.04,0.11] | −0.05 [−0.12,0.02] | −0.00 [−0.08,0.07] | ||
| 3MSE | |||||
| ILI | −0.09 (0.04) | −0.18 (0.04) | −0.13 (0.04) | 0.27 | 0.16 |
| DSE | −0.16 (0.04) | −0.10 (0.04) | −0.15 (0.04) | 0.53 | |
| Mean Difference | |||||
| [95% CI] | 0.07 [−0.03,0.18] | −0.07 [−0.18,0.04] | 0.02 [−0.09,0.13] | ||
| DSC | |||||
| ILI | −0.11 (0.03) | −0.26 (0.03) | −0.37 (0.04) | <0.001 | 0.15 |
| DSE | −0.15 (0.03) | −0.18 (0.03) | −0.39 (0.04) | <0.001 | |
| Mean Difference | |||||
| [95% CI] | 0.04 [−0.05,0.13] | −0.08 [−0.18,0.01] | 0.01 [−0.08,0.11] | ||
| Stroop | |||||
| ILI | −0.10 (0.04) | −0.24 (0.04) | −0.28 (0.04) | 0.004 | 0.36 |
| DSE | −0.10 (0.04) | −0.15 (0.04) | −0.31 (0.04) | 0.001 | |
| Mean Difference | |||||
| [95% CI] | 0.00 [−0.11..0.11] | −0.09 [−0.14,0.07] | 0.03 [−0.09,0.15] | ||
| TMT- A | |||||
| ILI | −0.08 (0.04) | −0.16 (0.04) | −0.21 (0.04) | 0.06 | 0.86 |
| DSE | −0.09 (0.04) | −0.13 (0.04) | −0.20 (0.04) | 0.12 | |
| Mean Difference | |||||
| [95% CI] | 0.01 [−0.10,0.12] | −0.03 [−0.14,0.07] | −0.01 [−0.12,0.11] | ||
| TMT-B | |||||
| ILI | −0.09 (0.04) | −0.19 (0.04) | −0.32 (0.04) | <0.001 | 0.53 |
| DSE | −0.10 (0.04) | −0.14 (0.04) | −0.24 (0.04) | 0.07 | |
| Mean Difference | |||||
| [95% CI] | 0.01 [−0.10,0.12] | −0.04 [−0.16,0.06] | −0.09 [−0.21,0.03] | ||
| RAVLT | |||||
| Immediate | |||||
| ILI | −0.13 (0.04) | −0.23 (0.04) | −0.18 (0.03) | 0.15 | 0.37 |
| DSE | −0.15 (0.04) | −0.16 (0.04) | −0.20 (0.04) | 0.72 | |
| Mean Difference | |||||
| [95% CI] | 0.02 [−0.07,0.12] | −0.06 [−0.17,0.03] | 0.02 [−0.09,0.13] | ||
| RAVLT | |||||
| Delayed | 0.71 | ||||
| ILI | 0.00 (0.04) | −0.07 (0.04) | −0.05 (0.04) | 0.36 | |
| DSE | −0.03 (0.04) | −0.05 (0.04) | −0.03 (0.04) | 0.90 | |
| Mean Difference | |||||
| [95% CI] | 0.02 [−0.09,0.13] | −0.02 [−0.13,0.08] | 0.03 [−0.07,0.13] |
With covariate adjustment for sex, age, education, race/ethnicity, and follow-up year.
95% confidence interval excludes 0.
Greater baseline FI was associated with poorer scores of executive function and processing speed (the DSC Stroop, and TMT-B tests), but less so for other domains including global cognition, verbal learning and memory. There was little evidence for differences between intervention groups for any cognitive domain: no interactions between intervention assignment and FI tertile reached statistical significance.
The baseline FI tertiles originally contained equal numbers of participants, however as seen in Table 1, there was relatively greater attrition by the time of cognitive assessments among those in the highest tertile compared to those in the first and second tertiles. This is evidence of differential lost follow-up, which has the potential to bias findings. To assess the sensitivity of results, we used inverse probability weighting to re-calculate Table 3 (Supplemental Exhibit S3). The mean cognitive function scores resulting from this approach tended to be lower than the unadjusted analyses, consistent with the expectation that participants resembling those lost to follow-up had lower cognitive functioning. However, when inverse probability weighting was used for inference, results were similar to unweighted analyses, which indicated that the differential attrition in the cohort did not materially biases results.
Within the ILI cohort, percent BMI changes (100% X (BMIYear8 − BMIBase)/BMIBase) from baseline to year 8 were unrelated to composite cognitive function: correlation coefficients ranged from 0.02 to 0.04 across the FI tertiles (Table 4). In contrast, for the DSE cohort, percent changes in BMI were inversely correlated with cognitive function for individuals in the second and third tertiles of FI, i.e. weight gain was associated with poorer cognitive function with 95% confidence intervals excluding 0 and differences in correlations among FI tertiles reaching p<0.01. Overall, patterns in percent change in BMI and cognitive among FI tertiles were significantly different by intervention groups (interaction p<0.01).
Table 4:
Correlation [95% confidence interval] between changes in body mass index from baseline to year 8 and composite cognitive function, by intervention assignment and baseline FI tertile.
| Baseline FI Tertile | Consistency of Relationships Among FI Tertile s: p-value | Interaction Between FI Tertile and Intervention Assignment: p-value | |||
|---|---|---|---|---|---|
| First | Second | Third | |||
| DSE | <0.01 | <0.01 | |||
| Correlation 95% CI | 0.02 [−0.00,0.04] | −0.06 [−0.08, −0.03] | −0.12 [−0.15,−0.09] | ||
| ILI | 0.37 | ||||
| Correlation 95% CI | 0.04 [0.02, 0.07] | 0.03 [0.00.0.05] | 0.02 [−0.01,0.05] | ||
4. Discussion
We report four findings: 1) higher baseline FI scores were associated with subsequent poorer cognitive functioning; 2) across the full range of FI, the ILI resulted in greater weight loss and increased physical fitness relative to DSE; 3) ILI, relative to DSE, was equally ineffective in benefiting cognitive function at all levels of FI; and 4) weight gain was increasingly associated with poorer cognitive function among individuals with higher baseline FI among DSE, but not ILI, participants.
4.1. Associations between FI and cognitive function
There are many reports that frailty, quantified via deficit accumulation, is associated with increased risk for cognitive decline and cognitive impairment.25,26 This is to be expected because often, and in our case, these indices contain many known risk factors for cognitive decline and there are many shared pathways between frailty and cognitive impairment, including obesity and insulin resistance.27 In Look AHEAD, higher FI scores were most strongly associated with relative deficits in measures of attention, processing speed, and executive function and these relationships were evident within both intervention groups. Others have reported that frailty is strongly associated with poorer and increased declines in processing speed28,29 and poorer and increased declines in executive function.29 Bunce et al., also found that the cognitive domains most strongly associated with FI were processing speed and executive function, and speculated that this is related to compromised circuity in in the frontal lobe.30 Unlike for the Look AHEAD cohort, however, increased frailty has also been reported to be associated with steeper deficits in memory.29
FI indices tended to increase over time as the participants grew older;23 the lack of a strong cross-sectional association between FI and age at baseline may reflect eligibility criteria and enrollment practices for this clinical trial cohort.
4.2. Consistency of intervention effects on weight
Participants were able to adhere to the intervention, i.e. lose meaningful amounts of weight and increase their physical fitness, across the full range of FI scores. We have previously reported that a range of age-related conditions, with the sole exception of self-reported worsening vision, did not appear to affect intervention adherence and the ability of individuals to sustain weight losses and increases in physical activity.31 Look AHEAD adopted a tailored approach to interventions, with treatment provided on an individual basis, which may have led to increased adherence across a range of health states.18
4.3. Lack of intervention effects on cognitive decline
We found no evidence that the Look AHEAD ILI had differential effects on cognitive function across the full range of the FI. This is contrary to our expectation that the intervention would benefit those with better age-related profiles. This expectation was based on the significant interactions that intervention assignment had with CVD history noted in the introduction.
While history of CVD and obesity were both associated with higher FI scores, this did not lead to a differential intervention effects on cognitive function related to FI. This suggests the above interactions are more specific to factors associated with CVD and obesity than with general biological aging. We have previously speculated that weight loss may lead to diminished neurovascular response to neurodegeneration,32 which potentially could be more tightly associated with CVD and obesity than with general aging.
We recently reported that among women, ILI was associated with relatively better cognitive functioning for those who began the intervention within 5 years of the menopause onset, but poorer cognitive functioning for those who began the intervention at least 10 years postmenopause.33 We repeated our analyses, stratifying by sex, but found no significant interactions between intervention and FI for either sex or any cognitive domain (data not shown).
We are curious why degree of weight loss among ILI participants, even those with high FI scores, was unrelated to cognitive function. One might expect that those with higher cognitive functioning, e.g. executive functioning, might be better able to adhere to the weight loss intervention, so that reverse causality might induce at least the appearance of a relationship. Among DSE participants, weight gain was increasingly associated with poorer cognitive function among those with elevated FI. DSE participants who gained weight, gained more than those in the ILI who gained weight, a mean (SD) for DSE of 6.1 (6.0) kg versus for ILI 5.4 (5.6), p=0.01. It may be that this contributed to a more pronounced relationship among DSE participants with higher baseline FI, however these differences in weight gains between intervention groups were similar across FI tertiles (interaction p=0.87). It is possible that ILI provided some protection that attenuated the relationship between weight gain and poorer cognitive function seen among DSE participants.
While the Look AHEAD ILI did not overall lead to long-term benefits in cognitive function, it provided many other important health benefits,34 some of which (e.g. hospitalizations, independent living) were more pronounced among older participants.35 We note that other multidomain lifestyle interventions that do not target weight loss have been shown to benefit cognition. For example, the Finnish Geriatric Intervention Study to Prevent Cognitive Impairment and Disability randomized controlled trial found that a 2 year multimodal lifestyle intervention consisting of nutritional guidance, exercise, cognitive training, and control of vascular and metabolic risk factors, when compared to general health advice, resulted in improved global cognition.36 This has led to the establishment of the World-Wide FINGERS international network to assess the efficacy of similar multidomain lifestyle interventions in different environments.37
4.4. Limitations
Our analyses are exploratory and post hoc and should be interpreted with caution. The Look AHEAD cohort, as comprised of volunteers for a clinical trial of a behavioral intervention who had diabetes and overweight/obesity, may not represent general clinical populations and may be, for the most part, too young to be undergoing marked cognitive decline. Rates of cognitive decline vary across cognitive tests,38 which may account for differences we see among cognitive domains. We lack cognitive assessments at trial baseline.
5. Conclusions
In a large cohort of adults with type 2 diabetes mellitus and overweight or obesity, an intensive lifestyle intervention to produce and sustain weight loss and increase physical activity had no long-term effect on cognitive function within any strata based on baseline deficit accumulation FI.
Supplementary Material
Key points.
It is unknown whether there is a window of opportunity when lifestyle interventions may slow cognitive aging.
Frailty indices based on the accumulation of health-related deficits are a marker of age-related health status.
Individuals with worse frailty index scores have lower subsequent cognitive function.
A 10-year intensive lifestyle intervention was equally ineffective in benefiting cognition across the full range of frailty.
ACKNOWLEDGMENTS
FUNDING
This research was funded by two diversity supplements to the Action for Health in Diabetes Extension Study Biostatistics Research Center (3U01DK057136-19S1 and 3U01DK057136-19S2). The funding for the parent award is from U01DK057136. Additional funding sources and the Look AHEAD study group are listed in Supplemental Exhibit S4.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Competing Interest Statement: The authors have no competing interests to disclose.
Clinicaltrials.gov Identifier: NCT00017953
REFERENCES
- 1.Chatterjee S, Peters SA, Woodward M, et al. Type 2 diabetes as a risk factor for dementia in women compared with men: a pooled analysis of 2.3 million people comprising more than 100,000 cases of dementia. Diabetes Care 2016;39:300–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Luchsinger JA, Gustafson DR. Adiposity, type 2 diabetes, and Alzheimer’s disease. J Alzheimers Dis 2009;16:693–704. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Espeland MA, Rapp SR. Bray GA, et al. Long-term impact of behavioral weight loss intervention on cognitive function: The Action for Health in Diabetes Movement and Memory Study. J Gerontol A Biol Sci Med Sci 2014;69;1101–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Rapp SR, Luchsinger JA, Baker LD, et al. Effect of a long-term intensive lifestyle intervention on cognitive function: Action for Health in Diabetes Study. J Am Geriatr Soc 2017;65:9966–9772. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Hayden KM, Baker LD, Bray G, et al. Long-term impact of lifestyle intervention on cognitive function assessed with the National Institutes of Health Toolbox: The Look AHEAD Study. Alz & Dement 2018;10:41–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Espeland MA, Luchsinger JA, Baker LD, et al. Effect of a long-term intensive lifestyle intervention on prevalence of cognitive impairment. Neurology 2017;88:2026–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Espeland MA, Dutton GR, Neiberg RH, et al. Impact of a multi-domain intensive lifestyle intervention on self-reported memory, problem-solving, and decision-making: The Action for Health in Diabetes Randomized Controlled Clinical Trial. J Gerontol A Biol Sci Med Sci 2018;73:1560–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Simpson FR, Pajewski NM, Beavers KM, et al. Does the impact of intensive lifestyle intervention on cardiovascular endpoints vary depending on age-related health status? J Gerontol A Biol Sci Med Sci 2020. June 20:glaa153. doi: 10.1093/gerona/glaa153. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Rockwood K, Mitnitski A. Frailty defined by deficit accumulation and geriatric medicine defined by frailty. Clin Geriatr Med 2011;27:17–26. [DOI] [PubMed] [Google Scholar]
- 10.Pajewski NM, Williamson JD, Applegate WB, et al. Characterizing frailty status in the Systolic Blood Pressure Intervention Trial. J Gerontol A Biol Sci Med Sci 2016;71:649–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Lee CMY, Woodward M, Batter ZGD, et al. Association of anthropometry and weight change with risk of dementia and its major subtypes: A meta-analysis consisting 2.8 million adults with 57 294 cases of dementia. Obes Rev. 2020. April; 21(4):e12989. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Hugenschmidt CE, Leng X, Lyles M, et al. Cognitive effects of adding caloric restriction to aerobic exercise training in older adults with obesity. Obesity 2019;27:1266–1274. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Napoli N, Shah K, Waters DL, Sinacore DR, Qualls C, Villareal DT. Effect of weight loss, exercise, or both on cognition and quality of life in obese older adults. Am J Clin Nutr 2014;100:189–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Luchsinger JA, Lehtisalo J, Lindström J, et al. Cognition in the Finnish diabetes prevention study. Diabetes Res Clin Pract. 2015;108(3):e63–6. [DOI] [PubMed] [Google Scholar]
- 15.Luchsinger JA, Ma Y, Christophi CA, et al. Metformin, lifestyle intervention, and cognition in the Diabetes Prevention Program Outcomes Study. Diabetes Care 2017;40:058–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.The Look AHEAD Research Group. Design and methods for a clinical trial of weight loss for the prevention of cardiovascular disease in type 2 diabetes. Control Clin Trials 2003;24:610–28. [DOI] [PubMed] [Google Scholar]
- 17.The Look AHEAD Research Group. Cardiovascular effects of intensive lifestyle intervention in type 2 diabetes. New Eng J Med 2013;369:145–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.The Look AHEAD Research Group. The Look AHEAD Study: a description of the lifestyle intervention and the evidence supporting it. Obesity 2006;14:737–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.The Look AHEAD Research Group. The development and description of the diabetes support and education (comparison group) intervention for the Action for Health in Diabetes (Look AHEAD) Trial. Clin Trials 2011;8:320–9.21730080 [Google Scholar]
- 20.Jakicic JM, Jaramillo SA, Balasubramanyam A, et al. Effect of a lifestyle intervention on change in cardiorespiratory fitness in adults with type 2 diabetes: results from the Look AHEAD Study. Int J Obes (London) 2009;33(3):305–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Espeland MA, Erickson K, Neiberg RH, et al. Brain and white matter hyperintensity volumes after ten years of random assignment to lifestyle intervention. Diabetes Care 2016;39:764–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Rapp SR, Luchsinger JA, Baker LD, et al. Effect of a long-term intensive lifestyle intervention on cognitive function: Action for Health in Diabetes Study. J Am Geriatr Soc 2017;65(5):966–972. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Simpson F, Pajewski NM, Nicklas B, et al. Impact of multi-domain lifestyle intervention on frailty through the lens of deficit accumulation in adults with Type 2 diabetes mellitus. J Gerontol A Biol Sci Med Sci 2020;75:1921–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Weuve J, Tchetgen EJ, Glymour MM, et al. Accounting for bias due to selective attrition: the example of smoking and cognitive decline. Epidemiology 2012;23:119–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Searle SD, Rockwood K. Frailty and the risk of cognitive impairment. Alz Res Therapy 2015;7:54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Rogers NT, Steptoe Cadar C. Frailty is an independent predictor of incident dementia. Scientific Rep 2017;7(1):15746. doi: 10.1038/s41598-017-16104-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Sargent L, Nalls M, Starkweather A, et al. Shared biological pathways for frailty and cognitive impairment: a systematic review. Ageing Res Rev 2018;47:149–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Rolfson DB, Wilcock G, Mitnitski A, et al. An assessment of neurocognitive speed in relation to frailty. Age Aging 2013;42:191–6. [DOI] [PubMed] [Google Scholar]
- 29.Thibeau S, McDermott K, McFail G, Rockwood K, Dixon RA. Frailty effects on non-demented cognitive trajectories are moderated by sex and Alzheimer’s genetic risk. Alz Res Therapy 2019;11:55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Bunce D, Batterham PJ, Mackinnon AJ. Long-term associations between physical frailty and performance in specific cognitive domains. J Gerontol B Psychol Sci Soc Sci 2019;74:919–26. [DOI] [PubMed] [Google Scholar]
- 31.Espeland MA, Rejeski WJ, West DS, et al. Intensive weight loss intervention in older individuals: Results from the Action for Health in Diabetes Type 2 Diabetes Mellitus Clinical Trial. J Am Geriatr Soc 2013:61:912–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Espeland MA, Luchsinger JA, Neiberg RH, et al. Long term impact of intensive lifestyle intervention on cerebral blood flow. J Am Geriatr Soc 2018;66:120–6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Yassine HN, Anderson A, Brinton R, et al. Do menopausal status and APOE4 genotype alter the long-term effects of intensive lifestyle intervention on cognitive function in women with type 2 diabetes mellitus? Neurobiol Aging 2020;92:61–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Pi-Sunyer X The Look AHEAD Trial: A review and discussion of its Outcomes. Curr Nutr Rep 2014;3(4):387–391. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Espeland MA, Glick HA, Bertoni AL et al. Impact of an intensive lifestyle intervention on use and costs of medical services among overweight and obese adults with type 2 diabetes: The Action for Health in Diabetes. Diabetes Care, 2014;37:2548–2556. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Ngandu T, Lehtisalo J, Solomon A, et al. A 2 year multidomain intervention of diet, exercise, cognitive training, and vascular risk monitoring versus control to prevent cognitive decline in at-risk elderly people (FINGER): a randomised controlled trial. Lancet. 2015;385(9984):2255–63. [DOI] [PubMed] [Google Scholar]
- 37.Kivipelto M, Mangialasche F, Snyder HM, et al. World-Wide FINGERS Network: A global approach to risk reduction and prevention of dementia. Alz Dement 2020;16(7):1078–1094. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Proust-Lima C, Amieva H, Dartigues JF, Jacqmin-Gadda H. Sensitivity of four psychometric tests to measure cognitive changes in brain aging-population-based studies. Am J Epidemiol. 2007;165:344–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.